What will Voter Intensity and Turn Out be in the 2014 Mid-Terms?

[HenryBHough]

Turnout will not follow norms this year. Not with Democrat thugs going house to house and forcing people to the polls.

 

[Judicial review]

Possibly but not likely.

Very likely.

 

[eagle1462010]

Low turn out will be the norm as always.............45% tops

 

[Judicial review]

Low turn out will be the norm as always.............45% tops

Over 50%.

 

[eagle1462010]

Low turn out will be the norm as always.............45% tops

Over 50%.

Too broad.......How much over 50%............within 5%.................

 

[Judicial review]

Low turn out will be the norm as always.............45% tops

Over 50%.

Too broad.......How much over 50%............within 5%.................

Doesn't matter. If the GOP gets over 50% it's all over.

 

[eagle1462010]

Low turn out will be the norm as always.............45% tops

Over 50%.

Too broad.......How much over 50%............within 5%.................

Doesn't matter. If the GOP gets over 50% it's all over.

Just predicting the turn out........Hope the turn out is higher..............

Either way, I want the GOP to take the Senate.

 

[WelfareQueen]

I think turnout will be around 45% nationally. Every early report I've seen says early voting is greatly exceeding expectations.

 

[Statistikhengst]

I hadn't had time to thank WelfareQueen for making this thread. Thanks, WQ.

I would like to add two points that are pretty neutral in every respect.

There is a difference between a partisan-breakdown model that some pollsters use to adjust their raw results, and the actual partisan self-identification of voters when asked during a survey.

In 2012, much, much, much polling showed a voter self-identification of D+7 to D+9, the mean being D+8. Many on the Right poo-pooed this, said it could not possibly be correct. Their bad. On election night, according to exit polls, the best estimate was between D+7 and D+8, pretty much right on the mark in terms of self-indentification during polling. I say "best estimate" because starting in 2012, not all states in the Union were exit-polled, which I consider to be a big mistake. It is possible that, had the exit polling included data from states like Idaho and Wyoming (which were not exit-polled), the margin would be smaller. Then again, Vermont and Rhode Island were not exit-polled, either, so maybe among the states that were not exit-polled, it was a zero-sum game. On election night, Obama won by +3.86%, or about 3.5 points under the partisan-self-identification spread, which is pretty on target, considering the Independent (unaffiliated) vote.

We had a similar situation in 2004, where much partisan self-identification was showing R+5 to R+6 in the electorate, and Ds were poo-pooing this as nonsense. Their bad. On election night, George W. Bush (43) won by +2.46%, about 3.5 points under the partisan-self-identification spread, which, like 2012 is pretty on target, considering the Independent (unaffiliated) vote.

The point I am making is that people should actually pay more attention to voter partisan SELF-identification statistics and try less to squeeze raw numbers into a formula that matches some upper-up's personal wishes.

This should lead some members to a question that I am deliberately not mentioning. Let's see who catches on.

Now, let's go to mid-term elections in terms of actual VT and let's take a look back in time:

Let's analyse the drop-off from a presidential election to the next mid-term:

1948 to 1950 (Truman's "second" term): -8

1952 to 1954 (Ike's 1st term): -18

1956 to 1958 (Ike's second term): -15

1960 to 1962 (Kennedy's only term): -16

1964 to 1966: (Johnson's only complete term): -14

1968 to 1970 (Nixon's 1st term): -15

1972 to 1974 (Nixon's second term, Ford took office on August 9, 1974, when all mid-term primaries were already over and done): -17

1976 to 1978 (Carter's only term): -16

1980 to 1982 (Reagan's 1st term): -12

1984 to 1986 (Reagan's 2nd term): -18

1988 to 1990 (Bush 41's only term): -14

1992 to 1994 (Clinton's 1st term): -19

1996 to 1998 (Clinton's 2nd term): -14

2000 to 2002 (Bush 43's 1st term): -14

2004 to 2006 (Bush 43's 2nd term): -20

2008 to 2010 (Obama's 1st term): -19

2012 to 2014 (Obama's 2nd term): ??

I bolded the statistic for mid-terms for all 2nd term-presidents (Truman 1950, Eisenhower 1958, Nixon/Ford 1974, Reagan 1986, Clinton 1998, Bush 43 2006). Let's look at them together:


1948 to 1950 (Truman's "second" term): -8

1956 to 1958 (Ike's second term): -15

1972 to 1974 (Nixon's second term, Ford took office on August 9, 1974, when all mid-term primaries were already over and done): -17

1984 to 1986 (Reagan's 2nd term): -18
1996 to 1998 (Clinton's 2nd term): -14

2004 to 2006 (Bush 43's 2nd term): -20

I increased the font size and purpled the values of president's who were quite unpopular in their second term of office, which is also the case with Pres. Obama at this time. We see values between -8 and -20.

I suggest that the Truman value is a major outlier, that the drop-off, if history is our guide, will be between -17 and -20. Median value: -18.5, rounded to -19.

That would place voter turnout in 2014 at 40%, which is what I predicted earlier. This is also in-line with 2010, 2006 and 2002, every mid-term election conducted after the beginning of the new millennium.

It may very well be true that more and more people are voting early. Those are the diehards (like myself) who want to make sure that their vote gets counted, no matter what. But that in no way guarantees high turnout on election day itself. I suspect that we will be between 39% and 41% VT, when the final canvasses are turned in and all is part of the congressional record.

Remember, election night statistics are far from complete, we usually have to wait up to 6, maybe 7 weeks, in order to get the final canvasses, signed, sealed and delivered.

Right now, we have about, I suspect, about 184 million registered voters. We had 183.1 million RV as of September 18, 2014, about 6 weeks ago. 40% of that total would be: 73.6 million ballots cast, but that can also be influenced by actual VT in big states compared to small states. I am doing end-of-October VR stats right now, this weekend; if this figure needs adjustment, then I will adjust accordingly.

WQ, I hope this information can be of help to you.

best,

Stat

CrusaderFrank
Derideo_Te
Nyvin
bendog
AceRothstein

 

[Statistikhengst]

Low turn out will be the norm as always.............45% tops

Over 50%.

Extremely unlikely.

 

[eagle1462010]

Mid terms are historically low turnout as the post from stat confirms. Higher turn out in the Presidential elections.

 

[Statistikhengst]

Mid terms are historically low turnout as the post from stat confirms. Higher turn out in the Presidential elections.

Yepp.

Plus, one or two percent of this is just a crap-shoot. If the weather is good, maybe a percent more. If it is bad (across party lines), one to two percent less. By this, I am referring to voting on election day itself.

 

[Derideo_Te]

Saw an interesting graphic on TRMS yesterday evening.

When there are "wave" election years, such as 2006 and 2010, the generic polling across all pollsters shows a strong trend towards whichever side is going to win usually by significant margins.

There is no such consistency going into the 2014 midterms. Generic polling is showing either +1 Dems or +4 Republican. Which seems to fit the pattern overall. My own interpretation is that in the red states there will be the usual RW turnout by the diehards and the Dems will hold their own in the Blue states but only by narrow margins.

When it comes to the closely contested races that is anyone's guess but conventional wisdom favors the Republicans while the Dems have a strong GOTV which I don't believe is showing in the polling. Personally I am expecting the GOP to take all of the close races so they end up with the majority but I am not ruling out a Dem surprise win somewhere or other which will restrict the Majority to only 51 or 52 tops.

If either of the 2 runoff states are involved then we won't know the final outcome until December or January. As of today this is the Republican's election to lose in my opinion.

 

[eagle1462010]

2014 Senate Election Simulation

This site is predicting the GOP to net gain 7 seats...............

The GOP will gain seats...........they need 6...............more conservative estimates put them at 50...........

21 seats up for grabs dem

15 seats up for grabs Gop

The GOP is pretty much a shoe in for 4 seats gained.........after that it's maybe.

 

[WelfareQueen]

I hadn't had time to thank WelfareQueen for making this thread. Thanks, WQ.

I would like to add two points that are pretty neutral in every respect.

There is a difference between a partisan-breakdown model that some pollsters use to adjust their raw results, and the actual partisan self-identification of voters when asked during a survey.

In 2012, much, much, much polling showed a voter self-identification of D+7 to D+9, the mean being D+8. Many on the Right poo-pooed this, said it could not possibly be correct. Their bad. On election night, according to exit polls, the best estimate was between D+7 and D+8, pretty much right on the mark in terms of self-indentification during polling. I say "best estimate" because starting in 2012, not all states in the Union were exit-polled, which I consider to be a big mistake. It is possible that, had the exit polling included data from states like Idaho and Wyoming (which were not exit-polled), the margin would be smaller. Then again, Vermont and Rhode Island were not exit-polled, either, so maybe among the states that were not exit-polled, it was a zero-sum game. On election night, Obama won by +3.86%, or about 3.5 points under the partisan-self-identification spread, which is pretty on target, considering the Independent (unaffiliated) vote.

We had a similar situation in 2004, where much partisan self-identification was showing R+5 to R+6 in the electorate, and Ds were poo-pooing this as nonsense. Their bad. On election night, George W. Bush (43) won by +2.46%, about 3.5 points under the partisan-self-identification spread, which, like 2012 is pretty on target, considering the Independent (unaffiliated) vote.

The point I am making is that people should actually pay more attention to voter partisan SELF-identification statistics and try less to squeeze raw numbers into a formula that matches some upper-up's personal wishes.

This should lead some members to a question that I am deliberately not mentioning. Let's see who catches on.

Now, let's go to mid-term elections in terms of actual VT and let's take a look back in time:

Let's analyse the drop-off from a presidential election to the next mid-term:

1948 to 1950 (Truman's "second" term): -8

1952 to 1954 (Ike's 1st term): -18

1956 to 1958 (Ike's second term): -15

1960 to 1962 (Kennedy's only term): -16

1964 to 1966: (Johnson's only complete term): -14

1968 to 1970 (Nixon's 1st term): -15

1972 to 1974 (Nixon's second term, Ford took office on August 9, 1974, when all mid-term primaries were already over and done): -17

1976 to 1978 (Carter's only term): -16

1980 to 1982 (Reagan's 1st term): -12

1984 to 1986 (Reagan's 2nd term): -18

1988 to 1990 (Bush 41's only term): -14

1992 to 1994 (Clinton's 1st term): -19

1996 to 1998 (Clinton's 2nd term): -14

2000 to 2002 (Bush 43's 1st term): -14

2004 to 2006 (Bush 43's 2nd term): -20

2008 to 2010 (Obama's 1st term): -19

2012 to 2014 (Obama's 2nd term): ??

I bolded the statistic for mid-terms for all 2nd term-presidents (Truman 1950, Eisenhower 1958, Nixon/Ford 1974, Reagan 1986, Clinton 1998, Bush 43 2006). Let's look at them together:


1948 to 1950 (Truman's "second" term): -8

1956 to 1958 (Ike's second term): -15

1972 to 1974 (Nixon's second term, Ford took office on August 9, 1974, when all mid-term primaries were already over and done): -17

1984 to 1986 (Reagan's 2nd term): -18
1996 to 1998 (Clinton's 2nd term): -14

2004 to 2006 (Bush 43's 2nd term): -20

I increased the font size and purpled the values of president's who were quite unpopular in their second term of office, which is also the case with Pres. Obama at this time. We see values between -8 and -20.

I suggest that the Truman value is a major outlier, that the drop-off, if history is our guide, will be between -17 and -20. Median value: -18.5, rounded to -19.

That would place voter turnout in 2014 at 40%, which is what I predicted earlier. This is also in-line with 2010, 2006 and 2002, every mid-term election conducted after the beginning of the new millennium.

It may very well be true that more and more people are voting early. Those are the diehards (like myself) who want to make sure that their vote gets counted, no matter what. But that in no way guarantees high turnout on election day itself. I suspect that we will be between 39% and 41% VT, when the final canvasses are turned in and all is part of the congressional record.

Remember, election night statistics are far from complete, we usually have to wait up to 6, maybe 7 weeks, in order to get the final canvasses, signed, sealed and delivered.

Right now, we have about, I suspect, about 184 million registered voters. We had 183.1 million RV as of September 18, 2014, about 6 weeks ago. 40% of that total would be: 73.6 million ballots cast, but that can also be influenced by actual VT in big states compared to small states. I am doing end-of-October VR stats right now, this weekend; if this figure needs adjustment, then I will adjust accordingly.

WQ, I hope this information can be of help to you.

best,

Stat

CrusaderFrank
Derideo_Te
Nyvin
bendog
AceRothstein

Great information Stat. So, what I can take from your information is that in the last 3 Presidential election cycles the actual end vote was roughly half of what the pre-polling party self-identification indicated.

Ex: Obama won 2012 with a 3.86% margin, but self-identification prior to the election revealed a +7 or + 8 margin. Ditto Bush in 2004 who won by 2.46% but pre-election self-identification was +5 to +6 Bush.

I know mid-terms are a different beast....but the data I have seen says anywhere from a +6 to +12 GOP advantage this cycle (see link in the OP). Assuming the Presidential model....the GOP gets roughly half of that advantage in actual votes nationally. That would put the GOP actual voter turnout advantage at about 4.5%.

If the general polling data is only factoring in a +2% to 2.3% GOP advantage that could make a significant difference in polling data around the Country.

Of course all of this is speculation....and mid-terms are more difficult to predict that Presidential Elections. But I think it is safe to say entities like Pew Research are probably correct. The GOP advantage this cycle will be at least equal to the 2010 mid-terms, and will probably exceed them.

 

[WelfareQueen]

For you Nate Silver fans......His breakdown. As most of you know...he was very accurate in 2012.


Senate Update With 4 Days Left Here 8217 s The State Of The Races FiveThirtyEight

 

[Derideo_Te]

The GOP advantage this cycle will be at least equal to the 2010 mid-terms, and will probably exceed them.

I disagree! 2010 was a wave year for the GOP. There is nothing to suggest that there is a similar wave in 2014.

The advantage lies with the GOP but they won't sweep to victory. It will be a nail biter with way too many races too close to call after the polls close and perhaps even a run off or two afterwards. Ultimately they will prevail in my opinion but it won't be a repeat of 2010.

 

[WelfareQueen]

The GOP advantage this cycle will be at least equal to the 2010 mid-terms, and will probably exceed them.

I disagree! 2010 was a wave year for the GOP. There is nothing to suggest that there is a similar wave in 2014.

The advantage lies with the GOP but they won't sweep to victory. It will be a nail biter with way too many races too close to call after the polls close and perhaps even a run off or two afterwards. Ultimately they will prevail in my opinion but it won't be a repeat of 2010.

De....you may be right. But Pew Research (I posted a link earlier in the thread) among other are seeing it very similar to 2010. We'll see....but I think most of the pollsters have it right.

 

[Derideo_Te]

The GOP advantage this cycle will be at least equal to the 2010 mid-terms, and will probably exceed them.

I disagree! 2010 was a wave year for the GOP. There is nothing to suggest that there is a similar wave in 2014.

The advantage lies with the GOP but they won't sweep to victory. It will be a nail biter with way too many races too close to call after the polls close and perhaps even a run off or two afterwards. Ultimately they will prevail in my opinion but it won't be a repeat of 2010.

De....you may be right. But Pew Research (I posted a link earlier in the thread) among other are seeing it very similar to 2010. We'll see....but I think most of the pollsters have it right.

I think they have hedged to be "right" on the final outcome but the odds of a wave are still far from certain because the Generic Polls are simply not showing the necessary "wave" uniformity.

This article gives a better insight into what I am seeing out there.

Races I 8217 ll Be Watching On Election Night

We won't know if there is going to be a wave until we see the returns on the early races. It might just end up as a GOP "ripple" instead.

 

[Statistikhengst]

Great information Stat. So, what I can take from your information is that in the last 3 Presidential election cycles the actual end vote was roughly half of what the pre-polling party self-identification indicated.

Ex: Obama won 2012 with a 3.86% margin, but self-identification prior to the election revealed a +7 or + 8 margin. Ditto Bush in 2004 who won by 2.46% but pre-election self-identification was +5 to +6 Bush.

I know mid-terms are a different beast....but the data I have seen says anywhere from a +6 to +12 GOP advantage this cycle (see link in the OP). Assuming the Presidential model....the GOP gets roughly half of that advantage in actual votes nationally. That would put the GOP actual voter turnout advantage at about 4.5%.

If the general polling data is only factoring in a +2% to 2.3% GOP advantage that could make a significant difference in polling data around the Country.

Of course all of this is speculation....and mid-terms are more difficult to predict that Presidential Elections. But I think it is safe to say entities like Pew Research are probably correct. The GOP advantage this cycle will be at least equal to the 2010 mid-terms, and will probably exceed them.

Not quite: the self-identification number usually only deals with D vs. R or R vs. D, but the real unknown is the unaffiliated factor, that is why at the end of the day, the national margin is lower than the D or R self-identification advantage. But 50% lower, not exactly.

As far as prognosticating for 2014, right now, in the Generic aggregate, it is only R +2.5, decidedly under where it was in 2010:

RealClearPolitics - Election Other - 2014 Generic Congressional Vote


In 2010, the end aggregate was GOP +9.3, which was more than three points to the Right of reality in that year. If we extrapolate that out to 2014, then if such errors in polling hold, then a generic aggregate of GOP +2.5 could, in reality, mean DEM +0.5 to DEM +1, but I consider that to be unlikely.

I screenshotted both of these data points and they are at the end of this posting:

2014 battle for control of the US Senate Page 10 US Message Board - Political Discussion Forum

Now, this is just one indicator and it is very possible that in some states, the Rs have a +12 to +15 advantage, but not nationally as a whole. I doubt that the GOP margin from 2010: GOP +5.7 (Senate) / +6.0 (House) will be exceeded. It is likely to be closer. No wave. Were a wave forming, signs of it would be appearing in the generic, and they are not there to be seen.

What the polling IS showing is that by and large, the GOP is winning in the Senatorial races where it needs to win. You don't even need a wave to do that, you just need pinpoint accuracy.